drivers/net/wireless/ath/ath10k/htt_tx.c

/*
 * Copyright (c) 2005-2011 Atheros Communications Inc.
 * Copyright (c) 2011-2013 Qualcomm Atheros, Inc.
 *
 * Permission to use, copy, modify, and/or distribute this software for any
 * purpose with or without fee is hereby granted, provided that the above
 * copyright notice and this permission notice appear in all copies.
 *
 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 */

#include <linux/etherdevice.h>
#include "htt.h"
#include "mac.h"
#include "hif.h"
#include "txrx.h"
#include "debug.h"

int ath10k_htt_tx_alloc_msdu_id(struct ath10k_htt *htt)
{
	struct ath10k *ar = htt->ar;
	int msdu_id;

	lockdep_assert_held(&htt->tx_msdu_pool_lock);

	if (htt->num_pending_tx >= htt->max_num_pending_tx)
		return -ENOBUFS;

	msdu_id = htt->free_msdu_ids[htt->free_msdu_ids_head];
	++htt->free_msdu_ids_head;
	if (htt->free_msdu_ids_head == htt->max_num_pending_tx)
		htt->free_msdu_ids_head = 0;

	++htt->num_pending_tx;
	if (htt->num_pending_tx == htt->max_num_pending_tx)
		ieee80211_stop_queues(htt->ar->hw);

	ath10k_dbg(ar, ATH10K_DBG_HTT, "htt tx alloc msdu_id %d\n", msdu_id);
	__set_bit(msdu_id, htt->used_msdu_ids);
	return msdu_id;
}

void ath10k_htt_tx_free_msdu_id(struct ath10k_htt *htt, u16 msdu_id)
{
	struct ath10k *ar = htt->ar;

	lockdep_assert_held(&htt->tx_msdu_pool_lock);

	if (!test_bit(msdu_id, htt->used_msdu_ids)) {
		ath10k_warn(ar, "trying to free unallocated msdu_id %d\n",
			    msdu_id);
	} else {
		htt->free_msdu_ids[htt->free_msdu_ids_tail] = msdu_id;
		++htt->free_msdu_ids_tail;
		if (htt->free_msdu_ids_tail == htt->max_num_pending_tx)
			htt->free_msdu_ids_tail = 0;

		--htt->num_pending_tx;
		if (htt->num_pending_tx == htt->max_num_pending_tx - 1)
			ieee80211_wake_queues(htt->ar->hw);
	}

	ath10k_dbg(ar, ATH10K_DBG_HTT, "htt tx free msdu_id %hu\n", msdu_id);
	__clear_bit(msdu_id, htt->used_msdu_ids);
}

int ath10k_htt_tx_alloc(struct ath10k_htt *htt)
{
	struct ath10k *ar = htt->ar;
	int tx_idx;
	spin_lock_init(&htt->tx_msdu_pool_lock);
	init_waitqueue_head(&htt->empty_tx_wq);

	if (test_bit(ATH10K_FW_FEATURE_WMI_10X, htt->ar->fw_features))
		htt->max_num_pending_tx = TARGET_10X_NUM_MSDU_DESC;
	else
		htt->max_num_pending_tx = TARGET_NUM_MSDU_DESC;

	ath10k_dbg(ar, ATH10K_DBG_BOOT, "htt tx max num pending tx %d\n",
		   htt->max_num_pending_tx);

	htt->pending_tx = kzalloc(sizeof(*htt->pending_tx) *
				  htt->max_num_pending_tx, GFP_KERNEL|__GFP_REPEAT|__GFP_HIGH);
	if (!htt->pending_tx)
		return -ENOMEM;

	htt->used_msdu_ids = kzalloc(sizeof(unsigned long) *
				     BITS_TO_LONGS(htt->max_num_pending_tx),
				     GFP_KERNEL|__GFP_REPEAT|__GFP_HIGH);
	if (!htt->used_msdu_ids) {
		kfree(htt->pending_tx);
		return -ENOMEM;
	}

	htt->tx_pool = dma_pool_create("ath10k htt tx pool", htt->ar->dev,
				       sizeof(struct ath10k_htt_txbuf), 4, 0);
	if (!htt->tx_pool) {
		kfree(htt->used_msdu_ids);
		kfree(htt->pending_tx);
		return -ENOMEM;
	}

	htt->free_msdu_ids = kmalloc(sizeof(int) *
					 htt->max_num_pending_tx,
					 GFP_KERNEL|__GFP_REPEAT|__GFP_HIGH);

	if (!htt->free_msdu_ids) {
		kfree(htt->used_msdu_ids);
		kfree(htt->pending_tx);
		dma_pool_destroy(htt->tx_pool);
		return -ENOMEM;
	}

	for (tx_idx = 0; tx_idx < htt->max_num_pending_tx; ++tx_idx) {
		htt->free_msdu_ids[tx_idx] = tx_idx;
	}
	htt->free_msdu_ids_head = 0;
	htt->free_msdu_ids_tail = 0;

	return 0;
}

static void ath10k_htt_tx_free_pending(struct ath10k_htt *htt)
{
	struct ath10k *ar = htt->ar;
	struct htt_tx_done tx_done = {0};
	int msdu_id;

	spin_lock_bh(&htt->tx_msdu_pool_lock);
	for (msdu_id = 0; msdu_id < htt->max_num_pending_tx; msdu_id++) {
		if (!test_bit(msdu_id, htt->used_msdu_ids))
			continue;

		ath10k_dbg(ar, ATH10K_DBG_HTT, "force cleanup msdu_id %hu\n",
			   msdu_id);

		tx_done.discard = 1;
		tx_done.msdu_id = msdu_id;

		// Unlocks tx_msdu_pool_lock since ath10k_txrx_tx_unref will acquire the lock again.
		spin_unlock_bh(&htt->tx_msdu_pool_lock);
		ath10k_txrx_tx_unref(htt, &tx_done);
		spin_lock_bh(&htt->tx_msdu_pool_lock);
	}
	spin_unlock_bh(&htt->tx_msdu_pool_lock);
}

void ath10k_htt_tx_free(struct ath10k_htt *htt)
{
	ath10k_htt_tx_free_pending(htt);
	kfree(htt->pending_tx);
	kfree(htt->used_msdu_ids);
	kfree(htt->free_msdu_ids);
	dma_pool_destroy(htt->tx_pool);
}

void ath10k_htt_htc_tx_complete(struct ath10k *ar, struct sk_buff *skb)
{
	dev_kfree_skb_any(skb);
}

int ath10k_htt_h2t_ver_req_msg(struct ath10k_htt *htt)
{
	struct ath10k *ar = htt->ar;
	struct sk_buff *skb;
	struct htt_cmd *cmd;
	int len = 0;
	int ret;

	len += sizeof(cmd->hdr);
	len += sizeof(cmd->ver_req);

	skb = ath10k_htc_alloc_skb(ar, len);
	if (!skb)
		return -ENOMEM;

	skb_put(skb, len);
	cmd = (struct htt_cmd *)skb->data;
	cmd->hdr.msg_type = HTT_H2T_MSG_TYPE_VERSION_REQ;

	ret = ath10k_htc_send(&htt->ar->htc, htt->eid, skb);
	if (ret) {
		dev_kfree_skb_any(skb);
		return ret;
	}

	return 0;
}

int ath10k_htt_h2t_stats_req(struct ath10k_htt *htt, u8 mask, u64 cookie)
{
	struct ath10k *ar = htt->ar;
	struct htt_stats_req *req;
	struct sk_buff *skb;
	struct htt_cmd *cmd;
	int len = 0, ret;

	len += sizeof(cmd->hdr);
	len += sizeof(cmd->stats_req);

	skb = ath10k_htc_alloc_skb(ar, len);
	if (!skb)
		return -ENOMEM;

	skb_put(skb, len);
	cmd = (struct htt_cmd *)skb->data;
	cmd->hdr.msg_type = HTT_H2T_MSG_TYPE_STATS_REQ;

	req = &cmd->stats_req;

	memset(req, 0, sizeof(*req));

	/* currently we support only max 8 bit masks so no need to worry
	 * about endian support */
	req->upload_types[0] = mask;
	req->reset_types[0] = mask;
	req->stat_type = HTT_STATS_REQ_CFG_STAT_TYPE_INVALID;
	req->cookie_lsb = cpu_to_le32(cookie & 0xffffffff);
	req->cookie_msb = cpu_to_le32((cookie & 0xffffffff00000000ULL) >> 32);

	ret = ath10k_htc_send(&htt->ar->htc, htt->eid, skb);
	if (ret) {
		ath10k_warn(ar, "failed to send htt type stats request: %d",
			    ret);
		dev_kfree_skb_any(skb);
		return ret;
	}

	return 0;
}

int ath10k_htt_send_rx_ring_cfg_ll(struct ath10k_htt *htt)
{
	struct ath10k *ar = htt->ar;
	struct sk_buff *skb;
	struct htt_cmd *cmd;
	struct htt_rx_ring_setup_ring *ring;
	const int num_rx_ring = 1;
	u16 flags;
	u32 fw_idx;
	int len;
	int ret;

	/*
	 * the HW expects the buffer to be an integral number of 4-byte
	 * "words"
	 */
	BUILD_BUG_ON(!IS_ALIGNED(HTT_RX_BUF_SIZE, 4));
	BUILD_BUG_ON((HTT_RX_BUF_SIZE & HTT_MAX_CACHE_LINE_SIZE_MASK) != 0);

	len = sizeof(cmd->hdr) + sizeof(cmd->rx_setup.hdr)
	    + (sizeof(*ring) * num_rx_ring);
	skb = ath10k_htc_alloc_skb(ar, len);
	if (!skb)
		return -ENOMEM;

	skb_put(skb, len);

	cmd = (struct htt_cmd *)skb->data;
	ring = &cmd->rx_setup.rings[0];

	cmd->hdr.msg_type = HTT_H2T_MSG_TYPE_RX_RING_CFG;
	cmd->rx_setup.hdr.num_rings = 1;

	/* FIXME: do we need all of this? */
	flags = 0;
	flags |= HTT_RX_RING_FLAGS_MAC80211_HDR;
	flags |= HTT_RX_RING_FLAGS_MSDU_PAYLOAD;
	flags |= HTT_RX_RING_FLAGS_PPDU_START;
	flags |= HTT_RX_RING_FLAGS_PPDU_END;
	flags |= HTT_RX_RING_FLAGS_MPDU_START;
	flags |= HTT_RX_RING_FLAGS_MPDU_END;
	flags |= HTT_RX_RING_FLAGS_MSDU_START;
	flags |= HTT_RX_RING_FLAGS_MSDU_END;
	flags |= HTT_RX_RING_FLAGS_RX_ATTENTION;
	flags |= HTT_RX_RING_FLAGS_FRAG_INFO;
	flags |= HTT_RX_RING_FLAGS_UNICAST_RX;
	flags |= HTT_RX_RING_FLAGS_MULTICAST_RX;
	flags |= HTT_RX_RING_FLAGS_CTRL_RX;
	flags |= HTT_RX_RING_FLAGS_MGMT_RX;
	flags |= HTT_RX_RING_FLAGS_NULL_RX;
	flags |= HTT_RX_RING_FLAGS_PHY_DATA_RX;

	fw_idx = __le32_to_cpu(*htt->rx_ring.alloc_idx.vaddr);

	ring->fw_idx_shadow_reg_paddr =
		__cpu_to_le32(htt->rx_ring.alloc_idx.paddr);
	ring->rx_ring_base_paddr = __cpu_to_le32(htt->rx_ring.base_paddr);
	ring->rx_ring_len = __cpu_to_le16(htt->rx_ring.size);
	ring->rx_ring_bufsize = __cpu_to_le16(HTT_RX_BUF_SIZE);
	ring->flags = __cpu_to_le16(flags);
	ring->fw_idx_init_val = __cpu_to_le16(fw_idx);

#define desc_offset(x) (offsetof(struct htt_rx_desc, x) / 4)

	ring->mac80211_hdr_offset = __cpu_to_le16(desc_offset(rx_hdr_status));
	ring->msdu_payload_offset = __cpu_to_le16(desc_offset(msdu_payload));
	ring->ppdu_start_offset = __cpu_to_le16(desc_offset(ppdu_start));
	ring->ppdu_end_offset = __cpu_to_le16(desc_offset(ppdu_end));
	ring->mpdu_start_offset = __cpu_to_le16(desc_offset(mpdu_start));
	ring->mpdu_end_offset = __cpu_to_le16(desc_offset(mpdu_end));
	ring->msdu_start_offset = __cpu_to_le16(desc_offset(msdu_start));
	ring->msdu_end_offset = __cpu_to_le16(desc_offset(msdu_end));
	ring->rx_attention_offset = __cpu_to_le16(desc_offset(attention));
	ring->frag_info_offset = __cpu_to_le16(desc_offset(frag_info));

#undef desc_offset

	ret = ath10k_htc_send(&htt->ar->htc, htt->eid, skb);
	if (ret) {
		dev_kfree_skb_any(skb);
		return ret;
	}

	return 0;
}

int ath10k_htt_h2t_aggr_cfg_msg(struct ath10k_htt *htt,
				u8 max_subfrms_ampdu,
				u8 max_subfrms_amsdu)
{
	struct ath10k *ar = htt->ar;
	struct htt_aggr_conf *aggr_conf;
	struct sk_buff *skb;
	struct htt_cmd *cmd;
	int len;
	int ret;

	/* Firmware defaults are: amsdu = 3 and ampdu = 64 */

	if (max_subfrms_ampdu == 0 || max_subfrms_ampdu > 64)
		return -EINVAL;

	if (max_subfrms_amsdu == 0 || max_subfrms_amsdu > 31)
		return -EINVAL;

	len = sizeof(cmd->hdr);
	len += sizeof(cmd->aggr_conf);

	skb = ath10k_htc_alloc_skb(ar, len);
	if (!skb)
		return -ENOMEM;

	skb_put(skb, len);
	cmd = (struct htt_cmd *)skb->data;
	cmd->hdr.msg_type = HTT_H2T_MSG_TYPE_AGGR_CFG;

	aggr_conf = &cmd->aggr_conf;
	aggr_conf->max_num_ampdu_subframes = max_subfrms_ampdu;
	aggr_conf->max_num_amsdu_subframes = max_subfrms_amsdu;

	ath10k_dbg(ar, ATH10K_DBG_HTT, "htt h2t aggr cfg msg amsdu %d ampdu %d",
		   aggr_conf->max_num_amsdu_subframes,
		   aggr_conf->max_num_ampdu_subframes);

	ret = ath10k_htc_send(&htt->ar->htc, htt->eid, skb);
	if (ret) {
		dev_kfree_skb_any(skb);
		return ret;
	}

	return 0;
}

int ath10k_htt_mgmt_tx(struct ath10k_htt *htt, struct sk_buff *msdu)
{
	struct ath10k *ar = htt->ar;
	struct device *dev = ar->dev;
	struct sk_buff *txdesc = NULL;
	struct htt_cmd *cmd;
	struct ath10k_skb_cb *skb_cb = ATH10K_SKB_CB(msdu);
	u8 vdev_id = skb_cb->vdev_id;
	int len = 0;
	int msdu_id = -1;
	int res;

	len += sizeof(cmd->hdr);
	len += sizeof(cmd->mgmt_tx);

	spin_lock_bh(&htt->tx_msdu_pool_lock);
	res = ath10k_htt_tx_alloc_msdu_id(htt);
	if (res < 0) {
		spin_unlock_bh(&htt->tx_msdu_pool_lock);
		goto err;
	}
	msdu_id = res;
	htt->pending_tx[msdu_id] = msdu;
	spin_unlock_bh(&htt->tx_msdu_pool_lock);

	txdesc = ath10k_htc_alloc_skb(ar, len);
	if (!txdesc) {
		res = -ENOMEM;
		goto err_free_msdu_id;
	}

	skb_cb->paddr = dma_map_single(dev, msdu->data, msdu->len,
				       DMA_TO_DEVICE);
	res = dma_mapping_error(dev, skb_cb->paddr);
	if (res)
		goto err_free_txdesc;

	skb_put(txdesc, len);
	cmd = (struct htt_cmd *)txdesc->data;
	cmd->hdr.msg_type         = HTT_H2T_MSG_TYPE_MGMT_TX;
	cmd->mgmt_tx.msdu_paddr = __cpu_to_le32(ATH10K_SKB_CB(msdu)->paddr);
	cmd->mgmt_tx.len        = __cpu_to_le32(msdu->len);
	cmd->mgmt_tx.desc_id    = __cpu_to_le32(msdu_id);
	cmd->mgmt_tx.vdev_id    = __cpu_to_le32(vdev_id);
	memcpy(cmd->mgmt_tx.hdr, msdu->data,
	       min_t(int, msdu->len, HTT_MGMT_FRM_HDR_DOWNLOAD_LEN));

	skb_cb->htt.txbuf = NULL;

	res = ath10k_htc_send(&htt->ar->htc, htt->eid, txdesc);
	if (res)
		goto err_unmap_msdu;

	return 0;

err_unmap_msdu:
	dma_unmap_single(dev, skb_cb->paddr, msdu->len, DMA_TO_DEVICE);
err_free_txdesc:
	dev_kfree_skb_any(txdesc);
err_free_msdu_id:
	spin_lock_bh(&htt->tx_msdu_pool_lock);
	htt->pending_tx[msdu_id] = NULL;
	ath10k_htt_tx_free_msdu_id(htt, msdu_id);
	spin_unlock_bh(&htt->tx_msdu_pool_lock);
err:
	return res;
}

int ath10k_htt_tx(struct ath10k_htt *htt, struct sk_buff *msdu)
{
	struct ath10k *ar = htt->ar;
	struct device *dev = ar->dev;
	struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)msdu->data;
	struct ath10k_skb_cb *skb_cb = ATH10K_SKB_CB(msdu);
	struct ath10k_hif_sg_item sg_items[2];
	struct htt_data_tx_desc_frag *frags;
	u8 vdev_id = skb_cb->vdev_id;
	u8 tid = skb_cb->htt.tid;
	int prefetch_len;
	int res;
	u8 flags0 = 0;
	u16 msdu_id, flags1 = 0;
	dma_addr_t paddr;
	u32 frags_paddr;
	bool use_frags;

	spin_lock_bh(&htt->tx_msdu_pool_lock);
	res = ath10k_htt_tx_alloc_msdu_id(htt);
	if (res < 0) {
		spin_unlock_bh(&htt->tx_msdu_pool_lock);
		goto err;
	}
	msdu_id = res;
	htt->pending_tx[msdu_id] = msdu;
	spin_unlock_bh(&htt->tx_msdu_pool_lock);

	prefetch_len = min(htt->prefetch_len, msdu->len);
	prefetch_len = roundup(prefetch_len, 4);

	/* Since HTT 3.0 there is no separate mgmt tx command. However in case
	 * of mgmt tx using TX_FRM there is not tx fragment list. Instead of tx
	 * fragment list host driver specifies directly frame pointer. */
	use_frags = htt->target_version_major < 3 ||
		    !ieee80211_is_mgmt(hdr->frame_control);

	skb_cb->htt.txbuf = dma_pool_alloc(htt->tx_pool, GFP_ATOMIC,
					   &paddr);
	if (!skb_cb->htt.txbuf)
		goto err_free_msdu_id;
	skb_cb->htt.txbuf_paddr = paddr;

	skb_cb->paddr = dma_map_single(dev, msdu->data, msdu->len,
				       DMA_TO_DEVICE);
	res = dma_mapping_error(dev, skb_cb->paddr);
	if (res)
		goto err_free_txbuf;

	if (likely(use_frags)) {
		frags = skb_cb->htt.txbuf->frags;

		frags[0].paddr = __cpu_to_le32(skb_cb->paddr);
		frags[0].len = __cpu_to_le32(msdu->len);
		frags[1].paddr = 0;
		frags[1].len = 0;

		flags0 |= SM(ATH10K_HW_TXRX_NATIVE_WIFI,
			     HTT_DATA_TX_DESC_FLAGS0_PKT_TYPE);

		frags_paddr = skb_cb->htt.txbuf_paddr;
	} else {
		flags0 |= SM(ATH10K_HW_TXRX_MGMT,
			     HTT_DATA_TX_DESC_FLAGS0_PKT_TYPE);

		frags_paddr = skb_cb->paddr;
	}

	/* Normally all commands go through HTC which manages tx credits for
	 * each endpoint and notifies when tx is completed.
	 *
	 * HTT endpoint is creditless so there's no need to care about HTC
	 * flags. In that case it is trivial to fill the HTC header here.
	 *
	 * MSDU transmission is considered completed upon HTT event. This
	 * implies no relevant resources can be freed until after the event is
	 * received. That's why HTC tx completion handler itself is ignored by
	 * setting NULL to transfer_context for all sg items.
	 *
	 * There is simply no point in pushing HTT TX_FRM through HTC tx path
	 * as it's a waste of resources. By bypassing HTC it is possible to
	 * avoid extra memory allocations, compress data structures and thus
	 * improve performance. */

	skb_cb->htt.txbuf->htc_hdr.eid = htt->eid;
	skb_cb->htt.txbuf->htc_hdr.len = __cpu_to_le16(
			sizeof(skb_cb->htt.txbuf->cmd_hdr) +
			sizeof(skb_cb->htt.txbuf->cmd_tx) +
			prefetch_len);
	skb_cb->htt.txbuf->htc_hdr.flags = 0;

	if (!ieee80211_has_protected(hdr->frame_control))
		flags0 |= HTT_DATA_TX_DESC_FLAGS0_NO_ENCRYPT;

	flags0 |= HTT_DATA_TX_DESC_FLAGS0_MAC_HDR_PRESENT;

	flags1 |= SM((u16)vdev_id, HTT_DATA_TX_DESC_FLAGS1_VDEV_ID);
	flags1 |= SM((u16)tid, HTT_DATA_TX_DESC_FLAGS1_EXT_TID);
	flags1 |= HTT_DATA_TX_DESC_FLAGS1_CKSUM_L3_OFFLOAD;
	flags1 |= HTT_DATA_TX_DESC_FLAGS1_CKSUM_L4_OFFLOAD;

	/* Prevent firmware from sending up tx inspection requests. There's
	 * nothing ath10k can do with frames requested for inspection so force
	 * it to simply rely a regular tx completion with discard status.
	 */
	flags1 |= HTT_DATA_TX_DESC_FLAGS1_POSTPONED;

	skb_cb->htt.txbuf->cmd_hdr.msg_type = HTT_H2T_MSG_TYPE_TX_FRM;
	skb_cb->htt.txbuf->cmd_tx.flags0 = flags0;
	skb_cb->htt.txbuf->cmd_tx.flags1 = __cpu_to_le16(flags1);
	skb_cb->htt.txbuf->cmd_tx.len = __cpu_to_le16(msdu->len);
	skb_cb->htt.txbuf->cmd_tx.id = __cpu_to_le16(msdu_id);
	skb_cb->htt.txbuf->cmd_tx.frags_paddr = __cpu_to_le32(frags_paddr);
	skb_cb->htt.txbuf->cmd_tx.peerid = __cpu_to_le32(HTT_INVALID_PEERID);

	ath10k_dbg(ar, ATH10K_DBG_HTT,
		   "htt tx flags0 %hhu flags1 %hu len %d id %hu frags_paddr %08x, msdu_paddr %08x vdev %hhu tid %hhu\n",
		   flags0, flags1, msdu->len, msdu_id, frags_paddr,
		   (u32)skb_cb->paddr, vdev_id, tid);
	ath10k_dbg_dump(ar, ATH10K_DBG_HTT_DUMP, NULL, "htt tx msdu: ",
			msdu->data, msdu->len);

	sg_items[0].transfer_id = 0;
	sg_items[0].transfer_context = NULL;
	sg_items[0].vaddr = &skb_cb->htt.txbuf->htc_hdr;
	sg_items[0].paddr = skb_cb->htt.txbuf_paddr +
			    sizeof(skb_cb->htt.txbuf->frags);
	sg_items[0].len = sizeof(skb_cb->htt.txbuf->htc_hdr) +
			  sizeof(skb_cb->htt.txbuf->cmd_hdr) +
			  sizeof(skb_cb->htt.txbuf->cmd_tx);

	sg_items[1].transfer_id = 0;
	sg_items[1].transfer_context = NULL;
	sg_items[1].vaddr = msdu->data;
	sg_items[1].paddr = skb_cb->paddr;
	sg_items[1].len = prefetch_len;

	res = ath10k_hif_tx_sg(htt->ar,
			       htt->ar->htc.endpoint[htt->eid].ul_pipe_id,
			       sg_items, ARRAY_SIZE(sg_items));
	if (res)
		goto err_unmap_msdu;

	return 0;

err_unmap_msdu:
	dma_unmap_single(dev, skb_cb->paddr, msdu->len, DMA_TO_DEVICE);
err_free_txbuf:
	dma_pool_free(htt->tx_pool,
		      skb_cb->htt.txbuf,
		      skb_cb->htt.txbuf_paddr);
err_free_msdu_id:
	spin_lock_bh(&htt->tx_msdu_pool_lock);
	htt->pending_tx[msdu_id] = NULL;
	ath10k_htt_tx_free_msdu_id(htt, msdu_id);
	spin_unlock_bh(&htt->tx_msdu_pool_lock);
err:
	return res;
}